1. Field of the Invention
The present invention relates to improving the read performance of a combined read-write head used for data storage on magnetic media. More particularly, the invention concerns reducing magnetic instability in the read head of a combined read-write head after writing, to improve read performance.
2. Description of the Related Art
Combined read and write heads, which are typically merged heads or piggyback heads, are commonly used for storing and retrieving data on magnetic media such as magnetic disks and tape. Combined read-write heads include both a read head and a write head. In a merged head, the read head and write head share a second shield that is also the first pole of the write head. In contrast, in a piggyback head, the second shield and the first write pole are separate parts separated by a magnetic insulating layer. The read head portion of a combined read-write head frequently employs a magnetoresistive (MR) sensor to read data. The resistance of a magnetoresistive sensor changes when the sensor is exposed to a magnetic field. A positive polarity magnetic field results in a different resistance than a negative polarity magnetic field. To read data, a current is passed through the MR sensor and the sensor is exposed to magnetic fields on a magnetic medium. The resulting voltages across the sensor, which are representative of the data on the magnetic medium, are then detected. Alternatively, a voltage can be applied to the sensor, and resulting changes in current through the sensor which are representative of the data on the magnetic medium are detected.
MR sensors are commonly designed with the magnetization of the head oriented at the optimal angle, also referred to as the optimal bias point, in relation to the direction of the current through the sensor. Exposure of the MR sensor to the relatively large magnetic fields of the pulses from the write head in a combined read-write head can disturb the domain state in the MR sensor, causing the response of the MR sensor, and thus the response of the head, to become nonlinear and/or inconsistent, and can also result in Barkhausen noise. Repeatability of the response of the MR sensor when the MR sensor is successively exposed to magnetic fields, and linearity of the response of the MR sensor when the MR sensor is exposed to a magnetic field, are key to proper operation of the read head. Thus, the nonlinearity and nonrepeatability induced by the write pulses is highly undesirable.
This magnetic instability in the MR sensor, which can be caused by the write pulses, can degrade the read performance of the head, resulting in errors when reading data. One commonly used measure of instability of read heads is the covariance of the head (which can also be referred to as the xe2x80x9cbias Aljohn numberxe2x80x9d). The covariance is a measure of the instability of the voltage amplitude of the read head after writing with the write head, and is given by dividing the standard deviation of the amplitude of the voltage across the MR sensor, by the mean amplitude of the voltage across the MR sensor. Other measures of magnetic instability include the data error rate, the servo error rate, and track misregistration. The data error rate is determined by writing data to the disk and then reading the data and measuring the number of errors. The servo error rate is the number of servo errors in a time period. Position error signals are monitored to determine the amount of track misregistration. For example, the standard deviation of the position error signals can be used as a measure of magnetic instability of the head.
FIG. 1A is a graph relating to the covariance of the voltage amplitude across the MR sensor of a first merged head. FIG. 1A depicts the amplitude of the voltage across the MR sensor at a variety of track positions. The head was tested on a spin-stand, with the polarity of the last write pulse before each read being random. 250 samples were taken at each track position, and the covariance of the voltage amplitude was calculated at each track position. The covariance values, which are indicative of the voltage amplitude instability of the MR sensor at each track position, are printed above the horizontal axis. For example, for this first head, the covariance at track position 1500 (on-track) is 2.3.
FIG. 2A is a similar graph depicting the amplitude of the voltage across the MR sensor of a second merged head. As with the first head, 250 samples were taken at a variety of track positions, with the polarity of the last write pulse before each read being random. Covariance values are printed above the horizontal axis. For this second head, the covariance at track position 1500 is 1.7.
Magnetic storage devices, such as disk drives, commonly employ a number of combined read-write heads, for example, 20 combined read-write heads. Magnetic instability in any of the heads is undesirable. Magnetic instability of heads adversely affects production yields of disk drives, degrades the performance of heads that pass initial testing, causes servo instability in sector servo files, and calls into question long-term magnetic stability during the life of the product. Accordingly, any improvement in the magnetic stability of heads is highly desirable.
Broadly, the present invention concerns a method, devices, and an article of manufacture for reducing magnetic instability in a read head of a combined read-write head, after writing data onto a magnetic storage medium.
The invention is based upon the discovery that, in a combined read-write head having a magnetoresistive (MR) read head, magnetic instability in the MR read head can be reduced or substantially eliminated if the last write pulse before each read always has the same designated polarity. The combined head can be any type of combined read-write head, and will typically be a merged head or a piggyback head. The designated polarity is either always positive or always negative, with one of the polarities usually producing better results than the other for a particular head. This is in contrast to the previous practice wherein the polarity of the last write pulse before a read is dependent upon the data written, and is therefore random. The polarity of last write pulses that results in the smallest amount of magnetic instability is determined empirically. One or more of a variety of methods for determining the amount of magnetic instability can be used. For example, the covariance of the voltage amplitude, the data error rate, the servo error rate, and/or the amount of track misregistration can all be used to determine the amount of magnetic instability after both positive polarity and negative polarity last write pulses.
To practice an illustrative embodiment of the invention, the last write pulse polarity that results in the least amount of magnetic instability for the head is determined, and is then referred to as the designated polarity. Next, after a set of write pulses is written, it is determined whether the last write pulse has the designated polarity. If the last write pulse does not have the designated polarity, then an additional write pulse with the designated polarity is written. Thus, the last write pulse before a read always has the designated polarity.
The invention affords its users with the distinct advantage of improving the magnetic stability of the MR head, which improves the linearity of the response when the MR head is exposed to a magnetic field, and improves the repeatability of the response of the MR head when the MR head is successively exposed to magnetic fields. Accordingly, the performance of the read head is improved. The present invention also provides other advantages and benefits, which are apparent from the following description.